Mesh bed leveling fails 3 out of 4 times  

RE: Mesh bed leveling fails 3 out of 4 times

I think some of the fails during mesh level are due to the printer detecting a spot that is too far out of "expected range" ... and making sure the printer is rock solid can help.  Trying placing the printer on the floor rather than a table; see if that helps.  If it does, you need to make your print stand stronger. Adding a brick or concrete stepping block under the printer has helped a few.

RE: Mesh bed leveling fails 3 out of 4 times

THanks for your replies. Unfortunately after securing the platform I am still having this problem. 

RE: Mesh bed leveling fails 3 out of 4 times

My MK3 had similar issues.  The PINDA was working perfectly for about a year.  Then it started failing occasionally during meshing.  The failures kept getting more frequent over a span of a few months.  I replaced the PINDA and everything was as good as new again.

Not sure if your problem is the same, but it wouldn't hurt to have a spare PINDA on hand.

Steve

RE: Mesh bed leveling fails 3 out of 4 times

I've got the same problem.  I looked for a post like this, but was in the MK3/MK3s forum thinking I was here.  I posted something there that a mod hopefully will paste to this thread.  EXACT same problem.

RE: Mesh bed leveling fails 3 out of 4 times

Scott, it looks like it is up to us to solve this.  I'll give you a description of my problem and what I've observed in trying to fix this.  Maybe we can find the answer between us.
Before the upgrade from MK2s to MK2.5s (I did a straight conversion, not bothering with the MK2.5 part), my printer was rock solid.  It would do mesh bed leveling quickly and without error.  Now I am constantly raising the extruder to the top of Z to attempt leveling again.  I can go anywhere from printing right away to about a dozen attempts before it prints.  Once the printer decides to print, everything is fine.  Failure can be on only point 8 to as bad as appearing indecisive at points 5, 6, and 8.  But it's always point 8 that stops the print.

I recall that when I replaced the non-magnetic printbed with the magnetic one, I had to work to get the last screw in place.  So I figured I may have forced a warp into the board by deforming it a bit to get that last screw to set.  So I took the bed off and expanded the holes in the frame with a 9/64" drill bit.  While the print bed was off I took a straightedge and a bright light to the surface of the bed to look for any warping.  There is a consistent warp from front to back down the center of the print bed, but it's very small.  This was checked with the spring steel sheet removed.  I should try again with the sheet in position, as it should average out any deformities.  I also tightened the bed back onto the frame fairly quickly, and in hindsight I think I should have let it sit on the frame with the fasteners loose so the board would be free to flex back into it's original form, if there is a chance of that happening.

My thinking is: If the printer eventually decides things are OK to print, then there is something moving, and I need to find out what it is.  OR, there is a problem with the firmware.  Since our problem appears to be small scale, I'm assuming our printers have something flexing that shouldn't.  Let's focus our efforts there.

Since the problem seems to be worse in the rear and left, I did a quick print job that ran the extruder diagonally from corner to corner and edge to edge, and watched the top of the screws for the Z axis.  I observed that as the extruder moved to the right it would be lowered.  I also noticed that the right threaded rod was moving quite a bit from front to back when the Z position changed.  Lastly, while looking for any clues at the top of the printer I saw that the smooth rods were extended up a little more on one side than the other.  To help with all this motion I reseated the smooth rods, checked the firmness in which all the bolts holding the Z-Axis motors and rods in place were and found they were not as tight as I would have liked.  Lastly, I went into setup and changed mesh bed leveling from 3x3 to 7x7, and the number of hops from 3 to 5.  This significantly increases the time before printing starts, but since I'm already nursing the printer anyway, I'm probably actually decreasing time before print until I find the problem.

I'm going to get a dial depth indicator and see if my bed really does have deformities.  I'm also going to loosen the screws holding the printbed in place and leave the bed hot for a while to see if I can undo any bend I may have introduced prior to opening the holes in the printbed frame.  If I find that my bed has a high corner, I'm going to get some extra nuts and mill them to height to get a true flatness from the top of the frame to the printbed.

So far, after tightening the Z-axis hardware and changing from 3x3 to 7x7 leveling I'm seeing more consistent print starts, but it is not doing the check as smoothly as my MK3 or itself prior to the upgrade.  It seems indecisive in the same areas where it would when doing a 3x3 check (4, 5, and 8).  As I explore this problem I'll keep you up to date.  Hopefully you will do the same?

RE: Mesh bed leveling fails 3 out of 4 times

Hi Scott,

I made some progress!  I loosened the screws holding the printbed to the frame, set the temperature of the heatbed to 60, and let it run where the bed could settle to whatever it wanted to until the timer turned the heater off.  Then I fastened everything together and did another Mesh Bed Level.  It was complaining, so I did it a few times, and then changed the settings back to 3x3.  Now I'm printing a new set of Z-Axis Top from the MK3s because that has less play for the rods to float around in, and for the first time since the upgrade, the MK2.5s shot through it's Pre Print Mesh Bed Leveling sequence just like it did before the upgrade.

Hopefully you can solve your issues by slightly enlarging the four holes in the Print Bed Frame (9/64" in my case), letting the print bed bend back to flat, and refastening everything.

In the meantime, I am going to replace as many of the old MK2 plastic parts as I can with the newer MK3 parts to decrease the amount of play/slop in my system.

RE: Mesh bed leveling fails 3 out of 4 times

I realize the Mk2 and Mk3 PINDAs are different, but many of are incorporating PINDA warmup strategies into our startup gcode. The idea is that the newer PINDA probes work best if mesh bed leveling is done at a consistent temperature (35C is recommended). While recent firmware has incorporated and possibly eliminated the need for the sometimes-lengthy PIND warmup, keeping the procedure in place also ensures that the print bed and surface are evenly warmed. If you can utilize the

M860 S35

gcode to wait for PINDA temps, might this help with some of these issues?

RE: Mesh bed leveling fails 3 out of 4 times

Thanks for the replies! Very interesting. I re printed the z axis mounts as well, but I have not yet done the z motor mounts. I will try that as well as maybe enlarging the holes. 

another thing to note. I have both a powder coated pei sheet that I use for printing PETG (at 230-240 degrees) and a spring steel sheet with smooth PEI that I use for printing PLA (at 210 degrees).  I notice the start up sequence is much better with the smooth sheet than the textured. Initially I thought it had something to do with the level of the sheets but Bobstro’s comment makes me thing it may have something to do with the bed temperature. I’m going to experiment with the bed temp. Settings and see if thag makes a difference. 

thanks again! 

RE: Mesh bed leveling fails 3 out of 4 times

@bobstro

Hi guys

I experimented with the startup sequence. The mesh bed leveling worked on the first try (amazing!!), but the bed did not warm to 35 degrees before the MBL sequence. Bobstro do you have any suggestions for edits to this sequence to work with your suggestions?

M862.3 P "[printer_model]" ; printer model check
M862.1 P[nozzle_diameter] ; nozzle diameter check
M115 U3.8.0 ; tell printer latest fw version
G90 ; use absolute coordinates
M83 ; extruder relative mode
G28 W ; home all without mesh bed level
M860 S35 ; set bed temp to 35
G80 ; mesh bed leveling
M104 S[first_layer_temperature] ; set extruder temp
M140 S[first_layer_bed_temperature] ; set bed temp
M190 S[first_layer_bed_temperature] ; wait for bed temp
M109 S[first_layer_temperature] ; wait for extruder tempG1 Y-3.0 F1000.0 ; go outside print area
G92 E0.0
G1 X60.0 E9.0 F1000.0 ; intro line
G1 X100.0 E12.5 F1000.0 ; intro line
G92 E0.0

RE: Mesh bed leveling fails 3 out of 4 times

To be clear, the M860 S35 waits for the PINDA to reach 35C. So far as I know, this requires the newer PINDA style with temp sensor, but I believe that's includes in the Mk2.5s upgrade. I heat the bed to 75C and hover the PINDA near it to speed the process along. It can take a while on the first print of the day, but you can watch the PINDA temp to verify. If the PINDA is already at or above 35C, it will move on immediately. Documented example here.

Due credit: I didn't originate this, only tweaked the process. Links to the relevant source pages are on my notes pages.

RE: Mesh bed leveling fails 3 out of 4 times

I am having the exact same issues at the same points.  Checking the level of the print bed, mine appears to be slightly higher in the very center where the two screws mount to the Y-axis carriage.  This is creating a slightly convex shape of the entire print bed. 

I attempted to remove the two screw at the rear of the print bed and slide a thin washer in between the carriage and the print bed mounting nut but still no change. It was still not quite enough of a spacer.  I am going to attempt to pick up 5mm longer screws tomorrow and try again with slightly thicker spacers on the rear and the thin ones up front to get this as flat as possible.

RE: Mesh bed leveling fails 3 out of 4 times

Guys,

At any stage did you check that both Y rods are fully seated in the z frame slots?

RE: Mesh bed leveling fails 3 out of 4 times

@billcampbell

Everything appears in place exactly as it should be.  Those components were never disassembled during the upgrade either.

RE: Mesh bed leveling fails 3 out of 4 times

SOLVED - This is a problem with the firmware. Issue begins with version 3.8.0 for MK2.5S and continues into 3.8.1.  Downgrading firmware to 3.7.2 instantly fixes the issue.  After a lot of troubleshooting and working with Prusa, I am back up and printing.  After all of this, the good news is also that my print bed is flatter than ever.  Love the upgrade over all.

Steps to fix:

- Reflash with firmware 3.7.2 (or any pre 3.8.0)

- Perform factory reset https://help.prusa3d.com/article/SYvbQ66IXF-factory-reset

- Calibrate via self test and calibration wizard

Happy printing

RE: Mesh bed leveling fails 3 out of 4 times

I just have same problem. Upgraded from mk2s to mk2.5s. New drivers/new firmware 3.8.1.

Self test all pass: XYZ calibration also ok.

I noticed first bug just at first attempt of first layer calibration. Printer was doing mesh bed leveling and stopped at middle point of rear side of bed. 

On screen message: Some problem encountered, Z-leveling enforced...

After this I need to turn knob to raise extruder axis all way up. After that printer repeats mesh bed leveling but again stops at same point.

After several repeated steps it would manage to complete first layer calibration, but problem was repeating on start of every new print. I had never issues like this on mk2s.

First contact over live chat and they told me to switch mesh bed pattern from 3x3 to 7x7, but that did not help. There was problem on same spot.

Second advice was to do PINDA temperature calibration. That also did not help.

Prusa support like had no more idea what was wrong.

I have also tried lowering pinda, testing pinda, checking for pinches but non of that helped.

Problem was also SOLVED by downgrading firmware to 3.7.2

There is only one thing I miss most and that is metal sheet profile. I have smooth PEI, and textured. Now I must do first layer calibration each time I switch sheet.

If problem started in 3.8.0 and continues to 3.8.1 I think Prusa is not doing any bug fixes. And new firmware is useless.

RE: Mesh bed leveling fails 3 out of 4 times

@marko-s4

The new firmware is more sensitive to height variations. The old firmware didn't care about that much.

Check carefully your sheet. Maybe there is a bump either on the sheet or on the bed which make the sheet bump up on that spot.

RE: Mesh bed leveling fails 3 out of 4 times

What was wrong with old firmware that they did have to change it to be more sensitive?

They could leave that part alone and just add new features. Now it is useless for me and some others.

Bed is new, and both sheets. I will tomorrow check for bed flatness at rear side but looking it by eye it looks ok. No bumps.

RE: Mesh bed leveling fails 3 out of 4 times

I am having the same problem. I followed the suggested fixes and went back to older firmware and it still stopped on the back row.  It has to be electron issues because sometimes it works properly and sometimes it doesn’t. I hope someone can figure this out.

RE: Mesh bed leveling fails 3 out of 4 times

Agreed and solved with firmware downgrade.

I also just recently upgraded my 3rd MK2 to the MK2S.  I didn’t have this problem with the first two upgrades and they are still running 3.8.1 (and have been for over 3 months).  My last upgrade wouldn’t get past point 8 (rear center) and 9/10 times I couldn’t even start a print after raising to top, it would still fail.

I downgraded to 3.7.2 and it hasn’t failed yet, on either 3x3 and 7x7 mbl.  

I DO miss being able to store the sheet profiles, as I have both types of sheets, but I simply have the number written down for each z height level that works on each sheet.

Hopefully they get this bug fixed with a later firmware update.  Thanks to the community for all their help with this.